Get first-hand advice on the intricate process of CMC development of protein therapeutics from this week’s guest contributor and one of the top voices in the field, Gene W. Lee.

CMC (Chemistry, Manufacturing, and Controls) is one of the key elements in drug development. And while CMC development is a simple catchphrase, it really encompasses a lot of different things, all of which need to be expertly executed and delivered at the right time.

Initially, the primary concern is centered around guaranteeing product safety, as well as confirming that the product can be feasibly manufactured without compromising its integrity or functionality.

Subsequently, the emphasis shifts to controlling the manufacturing process to ensure the efficacy, consistency, and quality of the recombinant product and the robustness of the process.

Oftentimes, in the eagerness to bring an exciting new molecule into the clinic , the start of CMC development happens too quickly, without all the information available to the CMC developer about the intent and purpose of the drug itself. So before we have a full knowledge of the target product profile, or TPP, or what the CMC process should look like, CMC development gets going. Why? Because CMC development is often on a critical path and takes a long time to get right. Unfortunately, missteps early on can lead to bigger problems later, resulting in escalating costs and project delays or cancellations.

As I mentioned above, CMC encompasses a lot of things, across many disciplines - vector construction, cell line development, cell culture and purification, analytical methods and characterization, formulation, GMP manufacturing and product release, primary and secondary packaging, and the list goes on. What doesn’t CMC do? Well, it can’t pump out drug product at a moment’s notice - this requires weeks or months of planning and preparation.  Manufacturing campaigns are expensive, and carefully planned to provide just enough material.  And it isn’t magic—if a protein is inherently prone to misfolding and aggregation, or clipping and heterogeneity, there’s only so much that the CMC team can do. But with increasing pressures to accelerate development timelines and go fast into the clinic, the expectation is that CMC will deliver products on time, and on budget. And so the pressure is to start early, with all the risks that come with an early start.

Start Early and Collaborate

The pressure to start early is to ensure readiness to have a CMC package to support your Investigational New Drug (IND) Application and get into the clinics as fast as possible. It’s starting work at risk, but it buys you a little bit of time as you continue to learn more about the product and the project.

So what are the consequences of that? I think for platform products, i.e., products that fit the platform process well, it is probably not such a big risk. You can start plugging your molecule into the platform process, trying to develop the most productive cell line, a suitable cell culture process, etc.  These activities take time to do, and that time can be spent learning more about the project.  Some of the key questions that the CMC team should seek to understand are:

• How much of the drug will you need to make to satisfy your Phase One requirements?
• How much of the drug will you need to satisfy your Phase Three or commercial requirements?
• What does that process need to look like?
• What are the costs of goods that will make your program competitive and successful?
• What is the formulation/concentration of your drug going to be?
• Are you planning to administer it as an IV or as a subcutaneous administration?

It is always a good idea to have conversations and discussions with as broad a team within the company or extended organization as possible. CMC shouldn't be working in isolation but rather within the context of fellow drug developers in the functional and therapeutic areas, as well as drug metabolism and pharmacokinetics (DMPK) scientists, pharmacologists, and clinical operations colleagues. The idea is to bring in as many bits of information as possible to know what you are trying to build, and to align around a common goal.

What About Platforms?

Platforms facilitate the rapid progression of a project to IND for several reasons, requiring minimal investment in process development, for instance, until clinical proof of concept. But why is it important to use a platform process?

If we take a corporate view, platforms enable shots on goal. Drug development is notoriously hard because most drugs that enter development will fail. One way to capitalize on novel targets or pathways and at the same time minimizing the loss on the investments that you are making in R&D is to have multiple shots on goal. And the more you can control costs, for example, through the use of platform processes, the more shots you can take. Platform processes have the additional benefit of speed, as there will be fewer surprises and a deeper experience than with bespoke processes.  This can translate to faster times for IND submission and entry into the clinic.  This approach fits in perfectly with the “phase appropriate” or “fit-for-purpose” philosophy of prioritizing speed to clinic in these early stages of CMC development.

I believe that in the field of biologics, we are extraordinarily lucky that monoclonal antibodies (mAb) have become the platform molecule of choice. They are easy to discover. They are easy to manipulate and engineer to bind to nearly any target you can imagine that is expressed on the cell surface or in circulation. We know that, of course, in the body, in the immune system, antibodies have evolved naturally through their combinatorial design to bind to tens of millions, if not more, unique targets.

From a drug discovery perspective, mAbs are a wonderful gift. And they also evolved to express themselves really well in B cells and plasma cells. And for this reason, as a drug discovery and drug development tool, it really fits in well with the concept of developing a platform process. I think naturally, as biotech companies and pharma companies adopted mAbs and developed these platforms, it became very easy for companies in the industry to compare how their platforms were working against other people’s platforms.

And then it becomes, of course, a little bit of a competition, a little bit of a race. You want to be able to produce your drug effectively and efficiently to reduce both the time and costs it takes to get your drug into the clinic, and over time new technologies are invented, developed, and implemented which enable better and more efficient platforms. Companies that have developed high performing and efficient antibody technologies can inspire others, further cementing mAbs as the platform of choice.

Why are Platforms Important?

Platforms are easy to implement, and it is easy to make monoclonal antibodies to hit a number of different targets. Companies can make the decision to try a lot of different possible candidates to go into the clinic—shots on goal. That is possible only by using a platform approach. However, it is important to consider that the industry is witnessing a lot of changes.

We are starting to see a shift away from mAbs. We have a lot of new modalities -  Fc-fusion proteins, bispecifics, T-cell engagers, novel scaffolds, and many more things are coming. At the same time, there is a sense of urgency because companies are running after the same, highly valuable targets.

The question is: To what extent does the platform approach still make sense today, or how should we adapt this approach?

Fortunately, a lot of these modalities are already sort of keeping platform approaches in mind in some way. In the case of Fc-fusion proteins, for example, the Fc part will bind to protein A and so fits into a purification platform quite well, at least at the critical initial capture step.  The same is true for bispecific antibodies.

By and large, the default starting point for expressing such molecules will still be CHO (Chinese Hamster Ovary) cells, together with a favorite cell culture medium and upstream process of choice. But, surprise!, not all antibodies express well in CHO or behave predictably in a platform purification process. And so even with a monoclonal antibody, the platform is not a guarantee of success, and the odds of success are worse with some of the novel modalities.  Can we improve our odds?

Developability Assessment

A great place to start is a developability assessment. If you have a good platform in place already, or if you are a small biotech firm and are working with a CDMO and they have a platform that you understand, you can try to see whether your design fits into that platform well. If it doesn’t, and if you have the opportunity to re-engineer your molecule or test different designs, you can test those in a developability assessment to see which ones work best on the platform that you ultimately want to use for manufacturing. That being said, of course, not all molecules will fit the platform. And this becomes more evident when you start straying from classical antibodies-like molecules.

For example, in the case of mRNA therapies or cell and gene therapies, platform rules (for biologics) don't apply anymore. Nevertheless, a lot of the thinking around process design and manufacturing has percolated into these areas as well. The industry as a whole has benefited from the 20 years or so of development of antibody platform processes.

How do Developability Assessments Help?

Historically, process development organizations would kick off CMC development work after lead candidate selection had taken place—the traditional hand-off between R&D and PD.  However, with CMC increasingly on the critical path, it is important to consider starting CMC development during candidate selection.  Typically, in drug discovery, you start with thousands, if not tens of thousands, of potential candidates in the hit discovery phase. So in an antibody screening campaign, for example, many candidates are screened to see which ones bind to your target with the highest affinity. That's a kind of functional screen that needs to be done in order to identify those antibodies that have the best potential to have the functional impact that you are looking for.

But as that screen begins to narrow down, you have the opportunity to look at other attributes of those antibody candidates—not just functionality, but now you can look at developability or manufacturability. Which of those antibodies has the potential to be expressed well in your large-scale bioreactor? Which of those antibodies has a tendency to aggregate under conditions that are representative of your cell culture process?

And while it might seem like you are now quite far away when you are working in a discovery lab from your 1,000 or 2,000-liter bioreactor, it is not too early to begin asking these questions.

A proper developability assessment can be as simple as expressing the top 20 candidates in a small-scale CHO transient system readily available in most discovery labs. And that will already give you some information about which of those antibodies has a better potential for good expression in a mammalian cell.

But a really sophisticated and more predictive developability assessment will use methods that are more representative of your larger-scale manufacturing process. It is better to express your antibodies not in a transient system, for example, but in a stable CHO system because we know that transients and stables don’t always agree with each other. And while small-scale expression is a good first step, it’s even better to perform an SEC analysis of your expressed protein to look at your high molecular weight profile because aggregation propensity at a small scale is quite predictive of how they will behave and perform at a larger scale. And then you can add additional layers.

How to Use Technology in Developability Assessment?

How can we leverage certain new technologies to overcome challenges linked to the greater complexity of molecules, such as  instability, fragmentation, aggregation, and a lot more?

One answer to that is to apply existing technologies as early as possible. In short, apply the technologies and methods that you are using today in process development labs to your developability assessment.

Those tools might need to be adapted for these early evaluations. One approach is miniaturization or scale-down versions of these tools when you are considering screening multiple candidates in a late discovery lab, perhaps in an automated multi-well format. But perhaps a more interesting and forward-thinking idea is to use AI assistance in bringing these complex modalities forward into development.

Prediction of a Protein’s 3D Structure

Programs like AlphaFold from DeepMind can accurately predict the 3D structure of any protein out there. As scientists design these new Frankenstein structures, putting pieces of interesting molecules together to build a candidate drug, algorithms such as AlphaFold or AI-assisted technologies can help us determine whether or not a structure is stable and can fold in a manner that will remain stable in solution.

With enough information and machine learning as well, we will come to a point where we may predict in silico many aspects of developability that we must determine experimentally at the moment.

And so the more layers you build into that early developability assessment, the better chance you will have of advancing candidates which will succeed through process development and manufacturing.

When starting early in CMC development, there's always a risk that you have placed your bets on the wrong horse, and it’s only after you have gone through this long development process that you find out that you have a three-legged horse. Or you have to expend a lot of effort and resources to optimize your process, which is now no longer strictly platform-based.

To avoid that, it is advisable to conduct a developability assessment before you have gone into real PD activities, before investing in a platform CMC development effort. Investing in these studies early can yield substantial results and considerable cost savings later.

How to Apply Developability Assessment?

The CMC-minded person on your early discovery project team must ask the question: Should we consider a structure or a molecule that has some developability aspects already built into it? Could we include an Fc portion in the molecule because we know that will be much easier to purify later on?

I think already, at the very beginning stages of protein design and engineering, one can begin introducing these developability concepts.

In the most basic form of a developability assessment, it is wise to look at productivity. You will need to make a minimum amount of protein, not just to serve the clinic but also for additional process development activities. You will also want to make sure that your molecule can be expressed with a minimum of protein aggregation.

The more questions you can build into your developability assessment, the more data you can pull out of that assessment, and the better decisions you will be able to make around your candidate. Developability assessments include, in addition to productivity and aggregation, hydrophobicity and a look at the pI of the antibody as well, which can have an influence on how to formulate your protein, and behavior under certain stress conditions as well.

How to Balance Investment in CMC Development?

In a large pharma company, a shot-on-target strategy can work well. Getting multiple candidates into late preclinical and then clinical development is a way to really de-risk your investment in R&D and have a chance of recouping that investment. It works very well if you have healthy portfolio development underpinning that strategy.

At a small biotech company, you are typically not working with a large portfolio of molecules. You are lucky if you have one or two molecules that you really believe in and that you want to bring all the way into commercialization. For such a firm, the platform approach is really a low-cost, low-investment way to get into the clinic very quickly and minimize your investment before clinical proof of concept.

As the CMC lead, I typically don’t like to overinvest in a process before clinical proof of concept. But if I don't have anything else coming up behind me in the portfolio, I then need to consider, well, does fast (and cheap) proof of concept make sense or does a more “all-in” strategy make sense? Because, really, the company’s success could very much depend on the success of one or two key assets in the clinic. It’s a different mindset, then.  “Phase appropriate” becomes meaningless if the strategy is to create a commercializable asset from the start.

In this case, I would recommend investing many more resources into CMC development before you have clinical proof of concept. For someone working at a smaller company, what would be the key parameters for success in CMC development, bringing your molecule into clinics, and finally, hopefully, into the market? Taking a look at the big picture is important here, understanding what you are trying to do with the drug and what the drug is ultimately designed to do, as well as what the company’s goals are.

Even for a difficult non-platform molecule, certain elements of that platform and of that CMC process will work fine without having to spend a lot of time on them, but other elements will be more problematic. For the project to succeed, for your CMC development to succeed, you need to make sure that all of those elements are successful. My advice is to identify those key areas and spend time there.

Another common challenge is the gap between R&D and process development, because these are different mindsets and different skill sets. Often in a smaller company, you don't necessarily have the manufacturing background on the team to guide CMC development. I suggest hiring a CMC expert or consultant because not all small biotech companies have the CMC experience necessary to prepare for the many challenges ahead.

Better Communication Techniques

Let's assume you have hired a CMC expert for your biotech company. Communication is absolutely the key here, whether you are working with an external CDMO or if you are fortunate enough to have an internal manufacturing organization. For example, R&D personnel in manufacturing often sit in very different areas. They have different cost centers, or may be in different time zones as well. It’s important that the lines of communication are established because of these very challenges.

We talked a lot about developability assessments. But developability really only makes sense in the context of your manufacturing platform. If R&D doesn’t know what your manufacturing platform is, they can't develop a good developability assessment. It’s important for that to be well understood by both sides. Here’s what you can do:

# 1 Foster Interpersonal Communication

Get to know each other. R&D and manufacturing are quite different. But at the end of the day, you both want to be successful in bringing your drug to the clinic, and that's the common ground. Find that common ground, get excited about it, and align around objectives. It is equally important for individuals to begin spreading out this interpersonal relationship to their teams so that it's not just reliant on one or two people but rather teams and organizations.

# 2 Document Best Practices

This enables future generations to implement what you have worked hard to build.

# 3 Be Flexible

It is wise to recognize that things will change. Your R&D strategy may change, your manufacturing strategy may change, and your platforms might change. And whatever you have documented that works today may not be relevant tomorrow. You need to update and modify whatever approach works well today.

See The Bigger Picture

Try to see the big picture as much as you can. If you are a biotech scientist or process development engineer in the lab, you are very focused on the task at hand. But you can talk to people who are upstream of you and downstream of you to see what it is you are doing and how it influences or is influenced by those adjacent groups. And if you have the opportunity to do so, look more and more outside of your area of responsibility and your organization to see how your work fits into the big picture. It is crucial to get context and structure for what each of us can do. And it’s a great reminder that we are all part of the journey to bringing important new drugs to the patients who need them.

Key Takeaways:

1. Start Early: Learn how starting CMC development early, even without a complete understanding of your drug, can reduce risks and prepare you better for clinical milestones.
2. Collaborate: CMC development should not work in isolation but integrate insights from diverse teams for a comprehensive drug development strategy.
3. Adopt a Platform Approach: Understand the role of platform processes in speeding up drug development and how they facilitate multiple attempts for cost-efficient progression from discovery to clinical trials, particularly in biologics.

About Gene W. Lee

Gene is the Chief Technical Officer at AltruBio, a biotech company located in San Francisco, CA focusing on novel, first in class biologics for autoimmune diseases. Prior to joining AltruBio, Gene was the global head of the Protein and Cell Sciences department at EMD Serono/Merck KGaA, where he was responsible for cell line development, analytical methods development, USP/DSP platform fit and GLP tox supply. He has also been the head of cell-line development at Percivia, and started his career at Genetics Institute/Wyeth BioPharma, in the "pre-platform" era.

Gene is an artist at heart - a trait he owes to his mother who encouraged him to pursue his artistic talents.  It helps him to always look at the big picture and to find creative solutions.

Connect on LinkedIn

Website: AltruBio Inc.  

Want to hear more? Go to Smart Biotech Scientist Podcast for insights and resources from our guests on various topics. Do you wish to simplify your biologics drug development project? Contact Us